AM3352: 3.3V power rail leakage

Hi Sung,

There appears to be about 500mV to 700mV on the 3.3V supply before any of the other supplies are started. This is likely causing the leakage path from VDDS 1.8V to 3.3V. Explanation is as follows: VDDS needs to reach 1.8V and remain stable before any other supply may start. So in the oscilloscope capture before the circle area there is some voltage on 3.3V (from other IC on the board) so once the 1.8V supply is started this will cause a leakage path from 1.8V to 3.3V which you see in the circle. And the path now raises the voltage from the first level of 500mV-700mV to the next level of about 1.30V.

The 1.10V -1.30V level on 3.3V is indicative that the leakage is coming from 1.8V through the internal ESD diodes (0.5V-0.7V diode drop).

First thing is check all other ICs on the board which are connected to the SoC and find out if any IC is receiving 3.3V while the SoC is off. I expect you will find another IC sending power to the SoC through I/O data pins. This needs to be isolated.